Endomorphisms of finite commutative groupoids related with multilayer feedforward neural networks

A. V. Litavrin

Trudy Instituta matematiki i mehaniki, Trudy Instituta Matematiki i Mekhaniki UrO RAN, Tome 27 (2021) no. 1, pp. 130-145

Résumé

In this paper, we introduce commutative, but generally not associative, groupoids $\mathrm{AGS}(\mathcal{N})$ consisting of idempotents. The groupoid $ (\mathrm{AGS}(\mathcal{N}),+)$ is closely related to the multilayer feedforward neural networks $\mathcal{N}$ (hereinafter just a neural network). It turned out that in such neural networks, specifying a subnet of a fixed neural network is tantamount to specifying some special tuple composed of finite sets of neurons in the original network. All special tuples defining some subnet of the neural network $\mathcal{N}$ are contained in the set $\mathrm{AGS}(\mathcal{N})$. The rest of the tuples from $\mathrm{AGS}(\mathcal{N})$ also have a neural network interpretation. Thus, $\mathrm{AGS}(\mathcal{N})=F_1\cup F_2$, where $F_1$ is the set of tuples that induce subnets and $F_2$ is the set of other tuples. If two subnets of a neural network are specified, then two cases arise. In the first case, a new subnet can be obtained from these subnets by merging the sets of all neurons of these subnets. In the second case, such a merger is impossible due to neural network reasons. The operation $(+)$ for any tuples from $\mathrm{AGS}(\mathcal{N})$ returns a tuple that induces a subnet or returns a neutral element that does not induce subnets. In particular, if for two elements from $F_1$ the operation $(+)$ returns a neutral element, then the subnets induced by these elements cannot be combined into one subnet. For any two elements from $\mathrm{AGS}(\mathcal{N})$, the operation has a neural network interpretation. In this paper, we study the algebraic properties of the groupoids $\mathrm{AGS}(\mathcal{N})$ and construct some classes of endomorphisms of such groupoids. It is shown that every subnet $\mathcal{N}'$ of the net $\mathcal{N}$ defines a subgroupoid $T$ in the groupoid $\mathrm{AGS}(\mathcal{N})$ isomorphic to $\mathrm{AGS}(\mathcal{N}')$. It is proved that for every finite monoid $G$ there is a neural network $\mathcal{N}$ such that $G$ is isomorphically embeddable into the monoid of all endomorphisms $\mathrm {AGS}(\mathcal{N}))$. This statement is the main result of the work.

Détail

Mots-clés : groupoid endomorphism
Keywords: multilayer feedforward neural networks, multilayer neural network subnet.

Bibliographie

[1] Golovko V. A., Krasnoproshin V. V., Neirosetevye tekhnologii obrabotki dannykh, uch. posobie, Izd-vo Belarus. gos. un-ta, Minsk, 2017, 263 pp.

[2] Gorban A. N., “Obobschennaya approksimatsionnaya teorema i vychislitelnye vozmozhnosti neironnykh setei”, Sib. zhurn. vychisl. matematiki, 1:1 (1998), 11–24 | MR

[3] McSulloh W., Pitts W., “A logical calculus of the ideas immanent in nervous activity”, Bulletin Math. Biophysics, 1943, no. 5, 115–133 | MR

[4] Rozenblatt F., Printsipy neirodinamiki: perseptron i teoriya mekhanizmov mozga, per. s angl., Mir, Moskva, 1965, 478 pp.

[5] Litavrin A. V., “Endomorphisms of Some Groupoids of Order $k + k^2$”, Bulletin Irkutsk State University. Ser. Mathematics, 32 (2020), 64–78 | DOI | MR | Zbl

[6] Tsarkov O. I., “Endomorphisms of the semigroup $G_2(r)$ over partially ordered commutative rings without zero divisors and with 1/2”, J. Math. Sc., 201:4 (2014), 534–551, N Y | DOI | MR | Zbl

[7] Zhuchok Yu. V., “Endomorphism semigroups of some free products”, J. Math. Sc., 187:2 (2012), 146–152, N Y | DOI | MR | Zbl

[8] Slepovichev I. I., “Algebraicheskie svoistva abstraktnykh neironnykh setei”, Izv. Sarat. un-ta. Nov. ser. Matematika. Mekhanika. Informatika, 16:1 (2016), 96–103 | DOI | MR | Zbl

[9] Plotkin B.I., Gringlaz L.Ya., Gvaramiya A.A., Elementy algebraicheskoi teorii avtomatov, Vysshaya shkola, M., 1994, 192 pp.

[10] Molchanov V. A., Khvorostukhina E. V., “Ob abstraktnoi opredelyaemosti universalnykh gipergraficheskikh avtomatov polugruppami vkhodnykh signalov”, Chebysh. sb., 20:2 (2019), 259–272 | DOI | MR | Zbl

[11] Litinskii L. B., “O zadache dekompozitsii neironnoi seti na neskolko podsetei”, Mat. modelirovanie, 8:11 (1996), 119–127 | MR

[12] Timofeev A. V., Derin O. A., “Printsipy postroeniya ierarkhicheskikh neironnykh setei dlya analiza multiizobrazhenii”, Tr. SPIIRAN, 10 (2009), 160–166

[13] Kurosh A. G., Lektsii po obschei algebre, Izd. dom “Lan”, SPb., 2007, 560 pp. | MR